Demand for accurate information on surgical quality is at an all time high. Unfortunately, most existing measures of surgical quality are flawed. The candidate proposes to develop composite measures of surgical morbidity that overcome many of the limitations of existing approaches. The candidate, Dr. Justin B. Dimick is a general surgeon and health services researcher at the University of Michigan. This proposal will facilitate his development into an independent scientist focused on optimizing surgical quality measurement. He will supplement his previous research training with additional course work in advanced statistical methods and pursue a mentored research plan with real world applicability to the American College of Surgeons-National Surgical Quality Improvement Program (ACS-NSQIP). The research plan has two specific aims: Aim 1. To develop composite measures of performance based on surgical morbidity. Using ACS-NSQIP data, the candidate will develop empirically derived composite measures of surgical performance. These measures will combine all relevant information on hospital quality to generate procedure-specific estimates of "true" morbidity that optimally filter out statistical noise. These measures will be validated by determining the extent to which they explain hospital level variation and forecast subsequent outcomes. Aim 2. To establish the value of incorporating process of care variables into composite measures. The ACS-NSQIP has begun routinely collecting data on the process measures set forth in the Surgical Care Improvement Program (SCIP). Using this data, the candidate will determine whether the addition of processes of care to composite measures improves their ability to explain hospital level variation and forecast subsequent outcomes. Completion of the proposed research will move forward the science of quality measurement and facilitate the candidate's transition into an independent investigator. [unreadable] [unreadable] [unreadable]